The epithelial component of breast cancers shows a remarkable level of cellular heterogeneity. This phenomenon can be best visualized by the variations of estrogen receptor (ER) expression in luminal breast cancer, which is the most common breast cancer subtype. Clinically, only 1% of ER positive cells are sufficient to justify the use of anti-endocrine therapy. In addition, ER remains a crucial therapeutic target even in patients who develop recurrent or metastatic disease. Despite the clinical importance of ER, it remains unclear how individual tumors maintain a balance of ER positive and negative cells. It would be of great clinical use to define whether this phenomenon is host or disease specific, and if it can be influenced to favor a better treatment response. To date, most research and drug development efforts have focused on intrinsic factors leading to cellular heterogeneity in ER+ breast cancer. However, our data support a profound and rapid effect of the tumor microenvironment on ER expression in tumor cells. We have identified two subtypes of cancer- associated fibroblasts (CAFs) within breast cancer stroma, which are defined by CD146 expression. The CAF subtypes differ in their gene expression profiles, their influence on ER expression in tumor cells, and their ability to direct the response to anti-endocrine therapy. Based on our preliminary data we hypothesize that CD146 positive and negative CAFs differentially modify ER expression in breast cancer cells by signaling through members of Notch, ErbB and Ephrin receptor families and have a critical effect on tumor response to treatment. To test this hypothesis, we will use approaches that combine our genetically and functionally defined primary human breast CAFs with breast cancer cell lines and our established patient-derived xenograft (PDX) models of ER+ breast cancer. In Specific Aim 1, we will define the mechanism responsible for CD146 CAF subtype-directed changes in ER expression and response to estrogen in breast cancer cells. Specific aim 2 will determine the influence of tumor-CAF crosstalk on development of anti-endocrine resistance. In specific aim 3, we will assess fibroblast subtypes in cancer and normal breast tissue, and their influence on response to anti-endocrine therapy in patient samples. The proposed studies will enhance our understanding of treatment resistance of luminal, ER+ breast cancer. Our ultimate goal is to identify approaches to target the tumor-stroma interactions, reduce the risk of disease recurrence and improve outcomes for patients with breast cancer.